Faster Growth Enhances Low Carbon Fuel and Chemical Production Through Gas Fermentation

Abstract

Gas fermentation offers both fossil carbon-free sustainable production of fuels and chemicals and recycling of gaseous and solid waste using gas-fermenting microbes. Bioprocess development, systems-level analysis of biocatalyst metabolism, and engineering of cell factories are advancing the widespread deployment of the commercialised technology. Acetogens are particularly attractive biocatalysts but effects of the key physiological parameter-specific growth rate (μ)-on acetogen metabolism and the gas fermentation bioprocess have not been established yet. Here, we investigate the μ-dependent bioprocess performance of the model-acetogen Clostridium autoethanogenum in CO and syngas (CO + CO₂+H₂) grown chemostat cultures and assess systems-level metabolic responses using gas analysis, metabolomics, transcriptomics, and metabolic modelling. We were able to obtain steady-states up to μ ∼2.8 day^-1 (∼0.12 h^-1) and show that faster growth supports both higher yields and productivities for reduced by-products ethanol and 2,3-butanediol. Transcriptomics data revealed differential expression of 1,337 genes with increasing μ and suggest that C. autoethanogenum uses transcriptional regulation to a large extent for facilitating faster growth. Metabolic modelling showed significantly increased fluxes for faster growing cells that were, however, not accompanied by gene expression changes in key catabolic pathways for CO and H₂ metabolism. Cells thus seem to maintain sufficient "baseline" gene expression to rapidly respond to CO and H₂ availability without delays to kick-start metabolism. Our work advances understanding of transcriptional regulation in acetogens and shows that faster growth of the biocatalyst improves the gas fermentation bioprocess.

Keywords: Clostridium autoethanogenum; acetogen; carbon recycling; chemostat; gas fermentation; genome-scale metabolic modelling; metabolomics; transcriptomics.

FIGURE 1

Specific growth rate-dependent growth characteristics in CO (top) and syngas (bottom) grown C. autoethanogenum chemostats. (A) Biomass and by-product concentrations. (B) Specific by-product production rates. (C) Specific gas uptake and production rates. Data are average ± standard deviation between bio-replicates (see Table 1). Asterisk denotes values statistically not different according to t-test (p-value < 0.05). μ, specific growth rate; gDCW, gram of dry cell weight; EtOH, ethanol; 2,3-BDO, 2,3-butanediol; Ace, acetate; q, specific rate; q _CO2 , specific CO₂ production rate; q _CO and q _H2 , specific CO and H₂ uptake rates, respectively.

FIGURE 2

Specific growth rate-dependent carbon balances in CO (top) and syngas (bottom) grown C. autoethanogenum chemostats. Carbon recoveries were 110 ± 16%, 103 ± 6%, 92 ± 1% for CO and 122 ± 6%, 107 ± 3%, 101 ± 3% for syngas at μ = 1.0, 2.0, 2.8 day⁻¹, respectively. Carbon recoveries were normalised to 100% to have a fair comparison of carbon distributions between different conditions. Data are average ± standard deviation between bioreplicates (see Table 1). μ, specific growth rate; 2,3-BDO, 2,3-butanediol.

FIGURE 3

Specific growth rate-dependent central metabolism flux levels in CO and syngas grown C. autoethanogenum chemostats. See dashed inset for bar chart details. Fluxes (mmol/gDCW/h) are represented as average ± standard deviation between bioreplicates (see Table 1). Arrows show direction of calculated fluxes; red arrows denote uptake or secretion; dashed arrows denote a series of reactions. Cofactors used in the GEM iCLAU786 are shown. Flux into PEP from oxaloacetate and pyruvate is merged. gDCW, gram of dry cell weight; μ, specific growth rate; OAA, oxaloacetate; PEP, phosphoenolpyruvate; THF, tetrahydrofolate; 2 PG, 2-phosphoglycerate; 2,3-BDO, 2,3-butanediol. See SIM1‒23 in Supplementary Tables S8, S9 for data and co-factor abbreviations.

FIGURE 4

Specific growth rate-dependent transcriptome characteristics in CO- and syngas-grown C. autoethanogenum chemostats. (A) Multidimensional scaling (MDS) of bioreplicate transcript abundances (RPKM). (B) Hierarchical clustering of bioreplicate transcript abundances (Z-scores based on RPKMs). (C) Correlation of bioreplicate transcript abundances (RPKM) (D) Venn diagrams showing overlap of DEGs (number of DEGs for each comparison in brackets). RPKM, reads per kilobase of transcript per million mapped reads; DEG, differentially expressed gene (fold-change > 1.5 with q-value < 0.05); μ, specific growth rate. See Supplementary Tables S2–S5 for RPKM and DEG data, respectively.

FIGURE 5

Global transcriptome changes assessed through functional gene classifications in CO- and syngas-grown C. autoethanogenum chemostats. (A) COG classification of DEGs for μ = 2.8 vs. 1.0 day⁻¹ (see Supplementary Table S6 for COG descriptions). (B) Hierarchical clustering and GO enrichment of expression changes of all 1,337 DEGs. Six identified expression clusters are illustrated with the coloured dendrogram with three clusters showing GO enrichment (clusters 2, 3, and 5). (C) Genes and GO terms with tightly controlled expression with faster growth. All 245 genes showing the same quantitative expression trend on both gas mixes are shown with 13 genes within the enriched GO terms denoted by coloured lines (CO, dashed; syngas, solid). GO:0032787, monocarboxylic acid metabolic process. See (B) for other GO terms. Gene IDs are preceded with CAETHG_. COG, Cluster of Orthologous Groups; DEG, differentially expressed gene (fold-change > 1.5 with q-value < 0.05); μ, specific growth rate; GO, Gene Ontology; q-Fisher, FDR-corrected p-value of Fisher’s exact test; AveExp, average of bioreplicates log₂ counts per million mapped reads (CPM; see Ritchie et al., 2015). See Supplementary Tables S4, S5, S7 for DEG data and GO terms list, respectively.

FIGURE 6

Individual gene expression changes in CO- and syngas-grown C. autoethanogenum chemostats. (A) Volcano plots showing up-(red) and down-regulated (blue) transcripts for μ = 2.8 vs. 1.0 day⁻¹. Genes in (B) are indicated by gene ID. Grey lines denote DEG thresholds (fold-change > 1.5 with q-value < 0.05) (B) μ-dependent expression profiles for DEGs linked to activities specifically relevant to acetogens (CO, dashed; syngas, solid). DEG, differentially expressed gene (fold-change > 1.5 with q-value < 0.05); μ, specific growth rate; AveExp, average of bioreplicates log₂ counts per million mapped reads (CPM; see Ritchie et al., 2015). Gene IDs are preceded with CAETHG_. See Supplementary Tables S4, S5 for DEG data. Volcano plot built using Enhanced Volcano package in R (Blighe et al., 2018).